[gcompris-devel] New activity to enrich the vocabulary

Bruno Coudoin bruno.coudoin at gcompris.net
Sun May 19 09:32:30 UTC 2013


Le 15. 05. 13 10:51, Fòram na Gàidhlig a écrit :
> My partner in crime has finished our first draft of the Gaelic
> translation. We got some serious expansion on some of the strings, so we
> hope that this will fit into the GUI.
>
> The biggest expansion is:
>
> hug	a' glacadh a chèile gu teann		933%	28 letters
>
>
> And the longest word:
>
> kneel	tha e a' cromadh air a ghlùinean	640%	32 letters
>
>
> Sorry, but for some of the pictures a 1-word translation just doesn't
> work at all.
>
Hi,

I don't know which string it is. Can you find out and test is this fit. 
If it does not, it is generaly easy to change a little bit the user 
interface. I'll be glad to get a patch.

>
>> Hi,
>>
>> For GCompris I created a specific file to manage the translation of the
>> voices.
>>
>> It is in the voices branch under en/words/content.txt
>>
>> The format is a simple key value pair list of the form:
>> ate.ogg:ate
>>
>> I did it this way instead of relying on the po translation because the
>> ogg content must match perfectly the text. A po translator would have to
>> hear all the voices. Also since we don't have all the voices it would
>> require asking po translator to translate words that are not available
>> in the game for their language.
> I agree this has to be in a separate file from the main PO.
>
> We have been thinking, what about introducing text domains? Then we
> could split up the content to translate to make it easier for new
> localisers to find out what's more important. How about something like this:
>
> - gcompris: GUI strings, settings, about page, game titles & main
> descriptions etc. Technical stuff like translating the language's
> alphabet (gletters, wordsgame, hangman).
>
> - gcompris-computer: strings for computer activities
>
> - gcompris-discovery: strings for discovery activities
I don't see a real advantage. In general translator don't mind if this 
is an important or less important string.

> As long as we keep this new activity in a separate file, the PO format
> will work just as well and be more comfortable for localizers.
> Translators actually need to look at the pictures with the original text
> on the abecedarium site rather than listen to the sound files. We only
> used the sound files to list which entries to translate.
>
No I don't think the po file works here. It is not about finding a name 
for an image but also selecting the same text as the voice.

Bruno.





More information about the Gcompris-devel mailing list